Preparation

Requirements

• Ubuntu 16.04
• Anaconda with python=3.6
• tensorFlow=1.12
• cuda=9.0
• cudnn>=7.4
• others: pip install termcolor opencv-python toposort h5py easydict

Compile custom operators

sh init.sh

Datasets

Shape Classification on ModelNet40

You can download ModelNet40 for here (1.6 GB). Unzip and move (or link) it to data/ModelNet40/modelnet40_normal_resampled.

Part Segmentation on PartNet

You can download PartNet dataset from the ShapeNet official webpage (8.0 GB). Unzip and move (or link) it to data/PartNet/sem_seg_h5.

Scene Segmentation on S3DIS

You can download the S3DIS dataset from here (4.8 GB). You only need to download the file named Stanford3dDataset_v1.2.zip, unzip and move (or link) it to data/S3DIS/Stanford3dDataset_v1.2.

The file structure should look like:

<tf-code-root>
├── cfgs
│   ├── modelnet
│   ├── partnet
│   └── s3dis
├── data
│   ├── ModelNet40
│   │   └── modelnet40_normal_resampled
│   │       ├── modelnet10_shape_names.txt
│   │       ├── modelnet10_test.txt
│   │       ├── modelnet10_train.txt
│   │       ├── modelnet40_shape_names.txt
│   │       ├── modelnet40_test.txt
│   │       ├── modelnet40_train.txt
│   │       ├── airplane
│   │       ├── bathtub
│   │       └── ...
│   ├── PartNet
│   │   └── sem_seg_h5
│   │       ├── Bag-1
│   │       ├── Bed-1
│   │       ├── Bed-2
│   │       ├── Bed-3
│   │       ├── Bottle-1
│   │       ├── Bottle-3
│   │       └── ...
│   └── S3DIS
│       └── Stanford3dDataset_v1.2
│           ├── Area_1
│           ├── Area_2
│           ├── Area_3
│           ├── Area_4
│           ├── Area_5
│           └── Area_6
├── init.sh
├── datasets
├── function
├── models
├── ops
└── utils

Usage

Training

ModelNet

python function/train_evaluate_modelnet.py --cfg <config file>  \
    [--gpus <list of gpus>] [--log_dir <log directory>]

• <config file> is the yaml file that determines most experiment settings. Most config file are in the cfgs directory.
• <list of gpus> means the indexes of gpus you want to use for training, like 0, 0 1, 0 1 2 3.
• <log directory> is the directory that the log file, checkpoints will be saved, default is log.

PartNet

python function/train_evaluate_partnet.py --cfg <config file>  \
    [--gpus <list of gpus>] [--log_dir <log directory>]

S3DIS

python function/train_evaluate_s3dis.py --cfg <config file>  \
    [--gpus <list of gpus>] [--log_dir <log directory>]

Evaluating

ModelNet40

python function/train_evaluate_modelnet.py --cfg <config file> --load_path <checkpoint> \
    [--gpu <gpu>] [--log_dir <log directory>]

• <config file> is the yaml file that determines most experiment settings. Most config file are in the cfgs directory.
• <checkpoint> is the model checkpoint used for evaluating.
• <gpu> means which gpu you want to use for evaluating, note that we only use one gpu for evaluating.
• <log directory> is the directory that the log file, checkpoints will be saved, default is log_eval.

PartNet

python function/evaluate_partnet.py --cfg <config file> --load_path <checkpoint> \
    [--gpu <gpu>] [--log_dir <log directory>]

S3DIS

python function/evaluate_s3dis.py --cfg <config file> --load_path <checkpoint> \
    [--gpu <gpu>] [--log_dir <log directory>]

Models

Method	ModelNet40	S3DIS	PartNet (val/test)
Point-wise MLP	92.8	66.2	48.1/51.2
Pseudo Grid	93.0	65.9	50.8/53.0
Adapt Weights	93.0	66.5	50.1/53.5
PosPool	92.9	66.5	50.0/53.4
PosPool*	93.2	66.7	50.6/53.8